Open Access

Virtual reality environments are increasingly being used to encourage individuals to exercise more regularly, including as part of treatment in those with mental health or neurological disorders. The success of virtual environments likely depends on whether a sense of presence can be established, where participants become fully immersed in the virtual environment. Exposure to virtual environments is associated with physiological responses, including cortical activation changes. Whether the addition of a real exercise within a virtual environment alters sense of presence perception, or the accompanying physiological changes, is not known. In a randomized and controlled study design, trials of moderate-intensity exercise (i.e. self-paced cycling) and no-exercise (i.e. automatic propulsion) were performed within three levels of virtual environment exposure. Each trial was 5-min in duration and was followed by post-trial assessments of heart rate, perceived sense of presence, EEG, and mental state. Changes in psychological strain and physical state were generally mirrored by neural activation patterns. Furthermore these change indicated that exercise augments the demands of virtual environment exposures and this likely contributed to an enhanced sense of presence.

Recently, virtual environments (VEs) are suggested to encourage users to exercise regularly. The benefits of chronic exercise on cognitive performance are well documented in non-VE neurophysiological and behavioural studies. Based on event-related potentials (ERP) such as the N200 and P300, cognitive processing may be interpreted on a neuronal level. However, exercise-related neuroelectric adaptation in VE remains widely unclear and thus characterizes the primary aim of the present study. Twenty-two healthy participants performed active (moderate cycling exercise) and passive (no exercise) sessions in three VEs (control, front, surround), each generating a different sense of presence. Within sessions, conditions were randomly assigned, each lasting 5 min and including a choice reaction-time task to assess cognitive performance. According to the international 10:20 system, EEG with real-time triggered stimulus onset was recorded, and peaks of N200 and P300 components (amplitude, latency) were exported for analysis. Heart rate was recorded, and sense of presence assessed prior to and following each session and condition. Results revealed an increase in ERP amplitudes (N200: p < 0.001; P300: p < 0.001) and latencies (N200: p < 0.001) that were most pronounced over fronto-central and occipital electrode sites relative to an increased sense of presence (p < 0.001); however, ERP were not modulated by exercise (each p > 0.05). Hypothesized to mirror cognitive processing, decreases of cognitive performance's accuracy and reaction time failed significance. With respect to previous research, the present neuroelectric adaptation gives reason to believe in compensative neuronal resources that balance demanding cognitive processing in VE to avoid behavioural inefficiency.